Journal Issue

Civil Liberties, Democracy, and the Performance of Government Projects

International Monetary Fund. External Relations Dept.
Published Date:
January 1998
  • ShareShare
Show Summary Details

Civil Liberties, Democracy, and the Performance of Government Projects

E MPIRICAL analysis demonstrates that the extent of civil liberties in a country affects the performance of a government’s investment projects. This finding contributes to accumulating evidence that the degree to which citizens’ “voices” in the public sphere are repressed or are allowed to be “heard” has an important influence on whether the accountability necessary for government efficacy will be created. In what is perhaps a surprising contrast, there is no clear relationship between indicators of electoral politics or “democracy” and the performance of government investments.

What, how, and how well?

The interrelationship between governments and economic development and poverty reduction is enormously complex. Three questions are relevant to politicians, policymakers, and their advisors: what should governments do; how should government decisions be made; and how well will governments be able to carry out their choices? In a wide variety of situations, it may be that the how well question is much more important than the what, since how well a government spends its resources may be more important than how much or what it spends its money on.

This article reports on new evidence that links how government decisions are made and how well they perform. This is obviously a broad set of questions, which we narrow to the relationship between overall rankings of countries’ political characteristics—in particular, the degree of civil liberties and the extent of democracy —and the performance of the set of government investment projects financed by the World Bank. Two indicators of a project’s success were used: the project’s economic rate of return, which is determined after project completion, and a simple rating indicating whether the project accomplished its developmental objectives.

One might think that Bank researchers would investigate the success or failure of Bank projects to examine the Bank’s performance. Although evaluators at the Bank do use such indicators for that purpose, data on the success of Bank-financed projects also serve as indicators of borrower country performance, for three reasons.

First, the projects financed by the World Bank are undertaken by the borrowing country’s government, which is responsible for project implementation. Governments chose to undertake these projects (although all of them must meet the Bank’s internal criteria for financing). Hence comparisons of ex post success indicate how well governments carry out projects they chose, rather than judging what projects they chose.

Second, the data give a unique ranking of project success on a comparable basis for a large number of countries. Most countries rarely evaluate systematically their own investment projects, much less do so in a way that would permit their findings to be compared with those of evaluations done by other countries. While there is a great deal of imprecision in the art of project assessment, the rankings obtained are fairly reliable indicators of success or failure.

Third, since the World Bank is a multilateral institution with written guidelines, procedures, and consistent methodologies that apply to all member country borrowers, it is unlikely that differences between countries in the success of Bank-supported projects are mainly a consequence of inconsistent actions of the Bank.


The study on which this article is drawn (see reference) did not construct a new measure of civil liberties but instead relied on four different cross-national rankings. One, undertaken by Freedom House, ranks countries annually based on a checklist of 14 criteria, which include media free of censorship, open public discussion, freedom of assembly and demonstration, and personal social rights (for example, the right to travel and the right to own property). Another rating, undertaken by Charles Humana, ranked countries in 1985 on the extent to which they complied with the human rights enumerated in the International Covenant on Civil and Political Rights adopted in 1966 by the United Nations General Assembly. These included, among others, the right of peaceful assembly, freedom of opinion and expression, and the right and opportunity to take part in public affairs. Two other narrower measures—of “media pluralism” and of “freedom to organize”—were also used and gave results similar to the more general indicators.

The most important finding of the current study was a strong and consistent link between measures of the extent of civil liberties in a country and the performance of World Bank-supported projects. Table 1 gives the estimated relationship between civil liberties and both the economic rate of return and project success. The average economic rate of return in the sample was 16 percent. The estimates obtained imply that if a country were to move from having the worst civil liberties observed to the best, the economic rate of return on its projects would improve by between 8 and 20 percentage points. This would mean increasing the return on projects by at least 50 percent and perhaps more than doubling the average return. Similar results are found for the summary indicator of project success: improving the degree of civil liberties in a country from the worst civil liberties in the sample considered by the current study to the best would reduce the rate of project failure by almost half.

The effects of civil liberties show up in both the social and economic sectors. The economic rate of return is calculated for only those sectors in which the calculation of financial returns is possible (for example, industry, infrastructure, and agriculture). These are the “hard” sectors where one might have believed governance factors would be less relevant than in the social, or “soft,” sectors, such as education and’ health. The results obtained in the study show an interesting symmetry with previous empirical work, which has shown that poor macroeconomic policies affect not just returns in the “economic” sectors but also project performance in the “soft” sectors like education and health (Kaufmann and Wang, 1995). This study shows that poor project performance, even in the economic and “hard” sectors, is related to poor governance.

The relationship between project success and civil liberties is not only reflected in a correlation between two variables but also holds when indicators of macroeconomic policy, external shocks, and countries’ economic features are included in an econometric model and when allowances are made for sector and regional differences in returns. Other, more recent investigations have also found positive relationships between Bank project success and other, broader indicators of “governance” quality such as the “index of credibility” reported in the Bank’s World Development Report 1997, and an indicator of “institutional quality” used by the Bank’s Operations Evaluation Department. We prefer measures of civil liberties, since we found rankings of civil liberties had greater power to explain the performance of government projects than did other indicators and because civil liberties are likely to have an effect precisely by affecting intermediate measures of government performance like credibility and corruption.

In addition to investigating the effects of civil liberties on the success of Bank-supported projects, we examined the data in more detail to see what effects the extent of political liberties (or of the existence of democracy) had. Again, to guard against peculiarities or anomalies in any particular measure, we used a variety of indicators of democracy. Freedom House provides a ranking of political liberties based on criteria such as whether the chief authority or legislature was elected by a meaningful process, there are multiple political parties, and there is freedom from domination by the military. Another researcher, Alberto Alesina, created a subjective index of democracy that ranks countries from 1 (most democratic) to 3. The Institutional Reform and Informal Sector (IRIS) Center at the University of Maryland has also created an indicator for democracy. Interestingly, none of these indicators of democracy or of political liberties had any impact on project performance.

Moreover, the issue of whether civil liberties or democracy was the more important influence on project performance was investigated. Although the two are quite highly correlated, the data indicate that the estimated impact of civil liberties on project success is not merely a consequence of the fact that democracies have extensive civil liberties. Civil liberties seem to be a powerful force for improving performance even in less than fully democratic settings, while if there were to be improvement in democracy without increases in civil liberties (which would admittedly be unusual), this would have no impact on performance.

Table 1.Relationship between a country’s civil liberties and its investment project performance
Indicator of pioject successPercentage point Increase. In economic rale of return

resulting from en improvement in civil liberties
Percentage pnint decrease In probability of project

failure resulting from an Improvement in civil liberties
By one standard deviationFrom worst to bestBy one standard deviationFrom worst to best
Freedom House1.983.213.2
Source: Jonathan Isham. Daniel Kaufmann. and Lam H Pnichert, 1997. “Civil Liberties. Democracy, and the Performance of Government Projects,” World Bank Economic Review, Vol 11 (May). PP. 219-42.Note The reported effects are estimated from a large sample of projects, controlling lor older country-speak structural and policy characteristics as well as for variations in the individual sectors.
Source: Jonathan Isham. Daniel Kaufmann. and Lam H Pnichert, 1997. “Civil Liberties. Democracy, and the Performance of Government Projects,” World Bank Economic Review, Vol 11 (May). PP. 219-42.Note The reported effects are estimated from a large sample of projects, controlling lor older country-speak structural and policy characteristics as well as for variations in the individual sectors.
Table 2.Indicators of expressions of citizen discontent for countries with high, medium, and low economic rates of return
Economic rain ofAverage economicRegionalDeviations of each activity from

population-adjusted averages

(average is zero for each indicator)
return category 1rate of returndistributionRiotsDemonstrationsStrikes
High22.2South Asia: 3: East Asia: 32.410.303.19
Medium17.0Latin America and

the Caribbean: 5;


Africa: 2;

Middle East and

North Africa: 3;

South Asia: 1

Africa: 9;

Latin America and

the Caribbean: 2;

South Asia: 1
Source: Authors calculations

Determined by average rales of return classified by country for all countries with at least 10 proiects during 1974-87

Source: Authors calculations

Determined by average rales of return classified by country for all countries with at least 10 proiects during 1974-87


We feel the best way to understand this new evidence on the relationship between civil liberties and government performance is to view it within the context of the debates on the micromechanics of government efficacy. No government works perfectly, of course. Examples of government misbehavior are abundant, ranging from severe political repression and kleptocracy on a massive scale to more mundane corruption, malfeasance, and indifference. While paying attention to “government failure” is a useful antidote to an abstract and naive characterization of governments, focusing on government failure is unproductive, since simply having no government at all is neither possible nor desirable, and rapid economic development requires certain crucial functions to be carried out by governments. The pressing question is not government size per se but rather what are the social, political, and institutional conditions that create incentives for governments to do well. This question obviously has no one answer, but the current study’s results on civil liberties are in line with the findings of interrelated research that emphasizes the “citizens’ voice” as a key mechanism. A centralized, top-down, technocratic approach to governance has often failed to deliver the hoped-for results.

The research on the impact of the citizens’ voice comes from a variety of settings. Accumulated empirical work has shown that the success of irrigation and rural water supply projects is sharply affected by the degree of citizen participation in their design, construction, and operation. In one study of 121 donor-financed water projects, those that were highly participatory were successful 68 percent of the time, while the projects with little or no participation by the intended beneficiaries of the projects were highly successful only 10 percent of the time (Isham, Narayan, and Pritchett, 1995). Research into the relative performance of irrigation systems in two Asian countries by Robert Wade also emphasized the degree of local control over the irrigation officers as a key element in how well the irrigation systems performed. Robert Putnam’s investigation into the relative efficacy of Italian regional governments settled on the social relationships among the citizens as the key determinant of government performance—a finding that is consistent with the conclusions of recent research work done on schools in Tanzania and on public spending in the United States. Judith Tendler’s recent examination of the enormous gains in health achieved in the Brazilian state of Ceara, Good Governance in the Tropics, suggests that governments can change rapidly from being unsuccessful to being successful by capturing the synergies among dynamic leaders, motivated cadres of workers, and local involvement.

More broadly, there is also evidence that countries with more extensive civil liberties (controlling for other variables) have performed better in providing schooling to girls and reducing infant mortality (as noted in the Bank’s World Development Report 1991). Ongoing research also suggests that improved civil liberties in a country also reduce the incidence of corruption.

The connection of civil liberties and the expression of the citizens’ voice explains an otherwise deeply puzzling finding. If one examines the data for connections between a variety of indicators of sociopolitical instability and project returns, one finds that, surprisingly, countries with higher economic rates of return had a much greater incidence of riots, strikes, and demonstrations (adjusted for population size) than countries with low rates of return (Table 2). Given that expressions of social unrest of this type are often thought to have an entirely negative influence on investment and performance, this was puzzling. The current study discovered, however, that countries with limited civil liberties had very little expression of social discontent, while much more expression occurred in countries with more extensive civil liberties. Indeed, once the study controlled for civil liberties, the apparent positive impact of riots and strikes disappeared completely. There is, however, an indication that countries in which all modes of expressing discontent are repressed experience worse performance on investment projects than countries in which public expression of social tensions is tolerated.

At the same time, narrowly defined political democracy is neither necessary nor sufficient to allow full expression of the citizens’ voice. Some of the worst, most ineffectual, and corrupt governments in the world are functioning democracies (while many others are not, of course), and some well-functioning states are arguably not fully democratic. The issues of who exercises power in the day-to-day operations of a country’s government are too complex to be fully resolved merely by finding out whether or not elections have been held.

Although the current study was primarily designed to use the data on World Bank-financed projects to examine the relative efficacy of governments using a comparable measure, the fact that these were donor-financed projects does raise questions about the implications for donors. The first and obvious implication is that the Bank has not been able to “cocoon” its own projects from the overall domestic environment. Earlier work on the relationship between the returns on Bank-supported projects and the macroeconomic environment played a role in changing the once-prevalent view that projects and policies could be neatly separated and that the performance of projects could be buffered from national policies. The interrelationship between a country’s macroeconomic policies and the success of aid in increasing its economic growth rate has been reconfirmed. (See “Aid Spurs Growth—in a Sound Policy Environment” by Craig Burnside and David Dollar, Finance & Development, December 1997.) In the context of the current study, a similar logic applies to project success and government efficacy.

There are very good arguments, however, for development agencies to have policies that keep them out of domestic politics. The present results do not suggest otherwise. In this context, two aspects of the current study’s findings should be kept in mind. First, the most hotly contested, narrowly political dimensions of democracy were not found to determine project performance. Second, most of the difficult questions about sovereignty, conditionality, and what types of donor selectivity are appropriate have revolved around questions of what the government should do. Alternatively, donors could place further emphasis on the how well dimension of government performance. The efficacy of the government activities that development agencies finance and encourage is obviously central to the agencies’ mission and hence is a natural concern. To the extent that project performance is determined by whether the expression of citizens’ voice and their participation in governance are encouraged— whether broadly through open and free discussions (and criticisms) and the existence of a free press, or narrowly through participation in the design and implementation of individual projects—this becomes a legitimate area for dialogue in the design of the projects to be financed.


As a final caveat on a sensitive and difficult area of discourse, the argument should not be made that civil liberties are important only because they are instrumental to greater government efficacy, since the reasons for this go much, much deeper. Nonetheless, civil liberties, along with other forms of expression and incorporation of citizens’ voice, do appear to have an instrumental value for improving a country’s economic performance and for designing the mechanisms for delivering government services of all types, from roads to schools. These linkages need to be better understood and incorporated—in ways that will vary, perhaps enormously, from sector to sector—into development activities.

NEW from the International Monetary Fund

Banking Soundness and Monetary Policy: Issues and Experiences in the Global Economy

Edited by Charles Enoch and John H. Green

Banking sector problems have affected many IMF member countries, and measures to remedy these problems as well as to prevent their recurrence deeply concern central bankers and policymakers The papers and comments published in this volume were presented at the Seventh Seminar on Central Banking, sponsored by the IMF. The topics include banking soundness and the macroeconomy, prudential standards, the role of the central bank during banking crises, and bank restructuring and supervision.

US$29.50. Available in English, (paper). ISBN 1-55775-645-7. xii + 542 pp. 1997. Stock #BSMWEA



Publication Services • Box FD-198 • 700 19th Street, N.W. • Washington, D.C. 20431 U.S.A.

Telephone: (202) 623-7430 • Telefax: (202) 623-7201 • E-mail: • Internet:

All orders must be prepaid. American Express, MasterCard, and VISA credit cards accepted.

a U.S. national, is a Senior Economist in the World Bank’s Development Research Group. Daniel Kaufmann, a Chilean national, is a Lead Economist in the World Bank’s Development Research Group.

This article draws on Jonathan Isham, Daniel Kaufmann, and Lant Pritchett, 1997, “Civil Liberties, Democracy, and the Performance of Government Projects,” World Bank Economic Review, Vol 11 (May), pp. 219-42.

Suggestions for further reading:

    Jonathan IshamDeepa Narayan andLant H. Pritchett1995Does Participation Improve Performance? Establishing Causality with Subjective Data,World Bank Economic ReviewVol. 9 (May) pp. 175200.

    • Crossref
    • Search Google Scholar
    • Export Citation

    JonathanIsham andDaniel KaufmannforthcomingThe Forgotten Rationale for Policy Reform,Quarterly Journal of Economics.

    DanielKaufmann andYanWang1995Macroeconomic Policies in the Social SectorsWorld DevelopmentVol. 23 pp. 751-65.

    World BankWorld Development Report1991:The Challenge of Development (New York:Oxford University Press for the World Bank).

    World BankWorld Development Report 1997: The State in a Changing World (New York:Oxford University Press for the World Bank).

Other Resources Citing This Publication